Co-simulation Applied to Carbon Capture
Technologies
Yang Fei
Submitted in accordance with the requirements for the degree of Doctor of Philosophy
The University of Leeds
Energy Technology Innovation Initiative School of Chemical and Process Engineering
The candidate confirms that the work submitted is his own, except where work which has formed part of jointly authored publications has been included. The contribution of the candidate and the other authors to this work has been explicitly indicated below. The candidate confirms that appropriate credit has been given within the thesis where reference has been made to the work of others.
The work performed in Chapter 6 of this thesis has been published in the following publication:
Y. Fei, S. Black, J. Szuhánszki, L. Ma, D.B. Ingham, P.J. Stanger, M. Pourkashanian, Evaluation of the potential of retrofitting a coal power plant to oxy-firing using CFD and process co-simulation, Fuel Processing Technology, 131 (2015) 45-58.
I developed the reduced order models in order to link the CFD predictions to the process modelling and also developed the oxy-coal power plant process model. Sandy Black and Janos Szuhanszki provided valuable assistance in setting up the CFD models. Penelope Stange gave meaningful suggestions on the process power plant. Lin Ma, Derek Ingham and Mohammed Pourkashanian are supervisors who provided helpful guidance on the overall direction and innovation for the research.
This copy has been supplied on the understanding that it is copyright material and that no quotation from the thesis may be published without proper acknowledgement.
The right of Yang Fei to be identified as Author of this work has been asserted by him in accordance with the Copyright, Designs and Patents Act 1988.
Acknowledgements
I would like to pass my thanks to Dr. Lin Ma for leading me the right way to learn and understand the background of chemical engineering; you always discussed with me so patiently on every academic or daily issue and helped me find the right way to work out problems. I would like to express my gratitude to Prof. Derek Ingham whose profound knowledge and kind personality continuously inspired me and he was always ready to help me with professional advices and experience. My sincere thanks should also go to Dr. Kevin Hughes, who helped me with so many academic and software issues. Thank you Prof. Mohamed Pourkashanian for always encouraging me with a big smile and providing professional suggestions when I was worried about the simulation results or the research progress. Thank you Prof. Gale, for spending so much time and energy to offer me strong support after the ETII team’s move to Sheffield. Also I would like to thank Sandy Black and Alastair Clements for helping me with CFD simulations. Thank Janos Szuhánszki for providing me the experimental data for the 250 kWth combustion test facility in Chapter 4.
My best friend Wen-lei Luo in Leeds, I remember the first day when I came to Leeds and it was you that picked me up at the train station in the midnight. I would like to acknowledge the China Scholarship Council and ETII, University of Leeds, for providing me with the financial support to perform this research work. RWE npower is acknowledged for the MOPEDS validation data. EPSRC is acknowledged for the meaningful support.
Abstract
In the energy supply sector, coal will still remain as a dominate role in the foreseeable future because: it is comparatively cheap and widely distributed around the world and more importantly, carbon capture and storage (CCS) technologies make it possible to depend on coal with almost zero emission of carbon dioxide (CO2). CCS involves capturing and purifying CO2 from the emission source and then sequestering it safely and securely to avoid emission to the atmosphere. Both the post-combustion and the oxy-fuel technologies can be applied to existing power plants for CCS retrofit. Accurate prediction of the performance of a CCS plant plays an important role in reducing the technical risk of future integration of CCS with existing power plants. This research combines the fundamental computational fluid dynamics (CFD) and system process simulation technologies so that an efficient co-simulation strategy can be achieved.
A 250 kWth coal combustion facility combined with a CO2 post capture plant is taken to test the conception of the CFD and process co-simulation approach. The CFD models are employed to account for the combustion facility and the predicted results on the outlet gas compositions, temperatures and mass flow rates are used to generate reduced order models to linked to the model for the PACT CO2 post capture plant so that a pilot scale whole plant model is achieved and validations have been made where it is possible.
Afterwards, the a large scale conventional air-coal firing power plant is taken into investigation: the CFD models for the boiler and the process models for the whole plant have been developed. Further, the potential of retrofitting this power plant to oxy-firing is evaluated using a CFD and process co-simulation approach. The CFD techniques are employed to simulate the coal combustion and heat transfer to the furnace water walls and heat exchangers under air-firing and oxy-firing conditions. A set of reduced order models has been developed to link the CFD predictions to the whole plant
process model in order to simulate the performance of the power plant under different load and oxygen enrichment conditions in an efficient manner. Simulation results of this 500 MWe power plant indicate that it is possible to retrofit it to oxy-firing without affecting its overall performance. Further, the feasible range of oxygen enrichment for different power loads is identified to be between 25% and 27%. However, the peak temperature on the superheater platen 2 may increase in the oxy-coal mode at a high power load beyond 450 MWe.
Table of content
Acknowledgements ... ii
Abstract ... iii
Table of content... v
List of Tables ... viii
List of Figures ... xi
Nomenclature ... xv
Chapter 1. Introduction and Motivation ... 1
1.1 Energy consumption and the role of coal ... 1
1.2 Coal combustion and its impacts on the environment ... 4
1.2.1Coal combustion in conventional power stations... 4
1.2.2Impacts on the environment ... 6
1.3 Carbon capture technologies ... 8
1.3.1Pre-combustion ... 9
1.3.2Post-combustion ... 10
1.3.3Oxy-fuel combustion ... 11
1.4 Power generation system modelling ... 12
1.5 Research aims, novelties and scope of the thesis ... 14
1.5.1Research aims and novelties ... 14
1.5.1Scope of the thesis ... 15
Chapter 2. Literature Review ... 17
2.1 Coal combustion process modelling ... 17
2.1.1Evaporation and devolatilisation ... 18
2.1.2Volatile combustion ... 20
2.1.3Char combustion ... 22
2.1.4Pollutant formation ... 24
2.2 Heat transfer and turbulence ... 26
2.2.1Heat transfer ... 27
2.2.2Turbulence ... 30
2.3 Carbon capture process modelling ... 34
2.3.1Chemical absorption process modelling ... 35
2.3.2Oxy-coal combustion process modelling ... 41
2.4 Summary ... 44
Chapter 3. Experimental Facilities and Data ... 45
3.1 The 250 kWth Combustion Test Facility (CTF) ... 45
3.1.1Facility introduction ... 45
3.1.2Burner description ... 46
3.1.3Measurements ... 47
3.1.4Fuel specification ... 48
3.1.5Experimental settings ... 50
3.2 The PACT amine capture plant ... 52
3.3 Didcot-A power plant ... 54
3.3.1Configurations of the power plant ... 54
3.3.2Fuel specification ... 55
3.3.3Boiler description ... 56
3.3.4Boundary conditions and available data for the boiler ... 57
3.4 Summary ... 60
Chapter 4. Modelling and Simulation of a Pilot Scale CO2 Capture System ... 61
4.1 CFD modelling of the 250 kWth air-coal combustion test facility ... 62
4.1.1Numerical set-up ... 62
4.1.2Model validation ... 66
4.1.3Simulation results and reduced order models ... 69
4.2 Integrated CFD and process modelling of the PACT facility... 72
4.2.1The gCCS system modelling environment ... 72
4.2.2Model validation ... 74
4.2.3The integration of the reduced order models into the process modelling and model settings ... 82
4.2.4Simulation results of the PACT facility ... 82
4.3 Summary ... 85
Chapter 5. Modelling and Simulation of a Large-scale Power Plant ... 87
5.1 CFD modelling of the full-scale coal fired boiler ... 87
5.1.1Model settings ... 87
5.1.2Coal data and boundary conditions ... 90
5.1.3Air-coal results and validation ... 93
5.1.4Air-coal and oxy-coal results analysis ... 94
5.2 The power plant simulations ... 98
5.2.1Full plant description ... 99
5.2.2Model components for the power plant ... 102
5.2.3Air-coal firing results and validation ... 112
5.2.4Air-coal and oxy-coal firing results analysis ... 114
5.3 Conclusions and limitations ... 119
Chapter 6. Evaluation of the Potential of Retrofitting a Coal Power Plant to Oxy-firing Using CFD and Process Co-Simulation ... 122
6.1 Research background ... 122
6.2 Essential component models for the co-simulation of the whole plant ... 124
6.2.1The natural circulation model ... 125
6.2.2The radiative heat exchanger model ... 126
6.2.3The furnace model ... 127
6.3 The ROM development ... 128
6.3.1Kriging interpolation ... 128
6.3.2Design of experiments (DOE) for the ROM development . 130 6.3.3The obtained ROMs ... 137
6.3.4Validation of the ROMs ... 140
6.4 Model validation and discussions on the whole plant co-simulations ... 141
6.4.1Validation of the integrated CFD/process full plant model ... 141
6.4.2Results and discussions ... 142
6.5 Conclusions ... 149
Chapter 7. Summary and the recommended work for f ... 150
7.1 Summary ... 150
7.2 Future work ... 154
List of Tables
Table 3.1 The El-Cerrejon coal analysis. ... 49
Table 3.2 Parameters for Rosin-Rammler distribution. ... 49
Table 3.3 Operating conditions for the air-coal experiments. ... 50
Table 3.4 The parameters for the absorber and the stripper columns. ... 52
Table 3.5 Essential components and instructions for the full plant. .... 55
Table 3.6 The Pittsburgh 8 coal analysis ... 55
Table 3.7 Coal combustion properties of Pittsburgh 8. ... 56
Table 3.8 Flow split fractions and swirl angels of the burners. ... 57
Table 3.9 Swirl directions of the burners. ... 57
Table 3.10 Air-coal boundary conditions for the boiler at full load condition. ... 58
Table 3.11 Heat transfer to different heat exchangers of the boiler at full load condition for the air-coal case obtained from MOPEDS. ... 58
Table 3.12 The gas and steam temperatures of the main heat exchangers obtained by MOPEDS. ... 59
Table 3.13 The steam generation rate, pressure and steam pressure of the steam drum obtained by MOPEDS. ... 59
Table 3.14 The steam flows, pressures, temperatures and power outputs from the steam turbines obtained by MOPEDS. ... 59
Table 4.1 Sub-models used in the CFD modelling of the 250 kWth coal combustion facility. ... 65
Table 4.2 The predicted outlet mass fractions, temperatures and the mass flow rates of the flue gas at different thermal inputs. .... 70
Table 4.3 The components and mass fractions assumed in the flue gas. ... 71
Table 4.4 The test conditions of the absorber column in the tests 32 and 47. (The values in the brackets have been adjusted.)... 75
Table 4.5 The validation results for the test 32. (The values in the brackets have been adjusted.) ... 76
Table 4.6 The validation results for the test 47. (The values in the brackets have been adjusted.) ... 77
Table 4.7 The input conditions of the stripper column in the tests 32 and 47. ... 78
Table 4.8 The considered thermal inputs and the corresponding mass flow rate and temperature of the flue gas. ... 83
Table 4.9 The simulation results of the PACT pilot plant with a
MEA mass fraction of 30% and a CO2 capture ratio of 90%. ... 84
Table 4.10 The simulation results of the PACT pilot plant with a MEA mass fraction of 35% and a CO2 capture ratio of 90%. ... 84
Table 4.11 The simulation results of the PACT pilot plant with a MEA mass fraction of 40% a CO2 capture ratio of 90%... 85
Table 5.1 Sub-models used in the CFD modelling of the boiler. ... 89
Table 5.2 Pittsburgh 8 coal analysis. ... 91
Table 5.3 Operating parameters for the air and oxy-coal cases. ... 91
Table 5.4 Average steam temperatures in the tube banks. ... 92
Table 5.5 Boundary conditions of the oxidant gas at each burner inlet. ... 93
Table 5.6 Heat transfer from the in-house code and the prediction from CFD for the air-coal case in the full-scale utility boiler. ... 94
Table 5.7 Essential components and simple instructions for the full plant model. ... 100
Table 5.8 PI/PID controllers used in the full plant model. ... 101
Table 5.9 A comparison in the temperature predictions on the steam side from MOPEDS and the full plant model for the heat exchangers. ... 113
Table 5.10 A comparison in the temperature predictions on the gas side from MOPEDS and the full plant model for the heat exchangers. ... 113
Table 5.11 A comparison in the predictions of MOPEDS and the full plant model for the steam drum. ... 113
Table 5.12 A comparison in the temperature and pressure predictions of MOPEDS and the full plant model for steam turbines (values in brackets are the MOPEDS results). ... 114
Table 5.13 The operating conditions for the air-coal and oxy-coal cases. ... 115
Table 6.1 Operating conditions of the sampling points for the CFD simulations of the furnace. ... 132
Table 6.2 Boundary settings for the operating burners at 31.7kg/s coal input rate. ... 132
Table 6.3 Boundary settings for the operating burners at 36.7kg/s coal input rate. ... 133
Table 6.4 Boundary settings for the operating burners at 41.7kg/s coal input rate. ... 133
Table 6.5 Boundary settings for the operating burners at 46.7kg/s coal input rate. ... 134
Table 6.6 Boundary settings for the operating burners at 51.7kg/s coal input rate. ... 134 Table 6.7 Heat transfer and furnace exit temperature predictions
from the boiler CFD simulations. ... 136 Table 6.8 Coal feed rates and oxygen concentrations of the
validation cases. ... 140 Table 6.9 Comparisons of heat transfer and temperature
predictions between the CFD and ROMs. ... 140 Table 6.10 A comparison in the temperature predictions on the
steam side from MOPEDS and the full plant model for the
heat exchangers. ... 141 Table 6.11 A comparison in the temperature predictions on the
gas side from MOPEDS and the full plant model for the heat
exchangers. ... 142
Table 6.12 A comparison in the predictions of MOPEDS and the full plant model for the steam drum. ... 142
List of Figures
Figure 1.1 World energy consumption by fuel [3]. ... 2
Figure 1.2 Fuel used in electricity generation in the UK over the last 15 years [1]. ... 2
Figure 1.3 Schematic of a coal fired sub-critical power plant. ... 4
Figure 1.4 Coal burner in a furnace in a power station [6]. ... 5
Figure 1.5 Global CO2 emissions since 1900 [7]. ... 7
Figure 1.6 Average atmospheric CO2 concentration since 1900 [7]... 7
Figure 1.7 Sea level rise over the last 100 years [8]. ... 7
Figure 1.8 A simplified diagram for the pre-combustion process [16]. ... 9
Figure 1.9 A simplified block diagram for the post-combustion process [16]. ... 10
Figure 1.10 A simplified block diagram for the oxy-fuel combustion process [16]. ... 12
Figure 2.1 Schematic of the combustion process of a coal particle [32]. ... 17
Figure 2.2 A process flow diagram for CO2 capture using chemical absorption approach [117]. ... 36
Figure 2.3 Descriptions of reactive absorption models with different abilities to describe the mass transfer and reaction kinetics [125]. ... 38
Figure 3.1 Layout of the 250 kWth CTF and a CAD image of the furnace [143]. ... 45
Figure 3.2 Images of the Doosan Babcock 250 kWth coal burner [143]. (a) burner with the quarl; (b) disassembled view showing from top to bottom: damper for tertiary and secondary split, tertiary inner pipe, secondary inner pipe, primary inner pipe, gas pipe; (c) assembled burner before installation and (d) burner installed in the CTF. ... 46
Figure 3.3 Sketch of the near burner region of the combustion rig. ... 47
Figure 3.4 Images of the IFRF suction pyrometer showing the (a) rear view, and (b) front vi, showing radiation shield [143]. ... 48
Figure 3.5 Images of a Medtherm GTW-50-24-21 584 heat probe [143]. ... 48
Figure 3.6 The fitted Rosin-Rammler curve [143]. ... 50
Figure 3.7 Measured radiative heat flux values along the inner wall of the furnace. ... 51
Figure 3.8 Measured temperature along the centre line inside the
furnace. ... 51 Figure 3.9 Process flow diagram of the PACT amine capture plant
[146]. ... 53 Figure 3.10 Configurations of the packing inside the absorber and
stripper columns. ... 53 Figure 3.11 Layout of the Didcot-A power plant. ... 54 Figure 3.12 A CAD drawing of the boiler and its burner. ... 57 Figure 4.1 Burner and furnace (a) CAD drawings, (b) simplified full
3D mesh, and (c) simplified periodic mesh [143]. ... 63 Figure 4.2 A CAD cross sectional view of the 250 kWth Doosan
Babcock burner [143]. ... 63 Figure 4.3 A CAD drawing of a typical section of the furnace [143]. .... 64 Figure 4.4 A predicted temperature distribution inside the furnace. .... 67 Figure 4.5 A plot of the temperatures along the centreline. ... 68 Figure 4.6 A plot of the surface incident radiation along the wall. ... 68 Figure 4.7 The predicted temperature profiles in the furnace with
different thermal inputs. ... 70 Figure 4.8 The predicted velocity profiles in the furnace with
different thermal inputs. ... 70 Figure 4.9 Temperature of the flue gas as a function of thermal
input. ... 72 Figure 4.10 Mass flow rate of the flue gas as a function of thermal
input. ... 72 Figure 4.11 Schematic representation of the two-film theory [158]. ... 73 Figure 4.12 Absorber temperature measurement locations [167]... 74 Figure 4.13 The flow sheet of the absorber column generated in
gCCS. ... 76 Figure 4.14 The predicted temperatures along the height of the
column for the test 32. ... 77 Figure 4.15 The predicted temperatures along the height of the
column for the test 47. ... 78 Figure 4.16 The flow sheet of the stripper column generated in
gCCS. ... 79 Figure 4.17 The predicted temperatures along the height of the
column for the test 32. ... 79 Figure 4.18 The predicted temperatures along the height of the
column for the test 47. ... 80 Figure 4.19 A flowsheet for the whole CO2 capture process in
Figure 4.20 The predicted temperature profile in the absorber by
the standalone and the integrated models for the test 32. ... 81 Figure 4.21 The predicted temperature profile in the stripper by
the standalone and the integrated models for the test 32. ... 81 Figure 4.22 A flowsheet for the PACT amine plant generated in
gCCS. ... 82 Figure 5.1 CFD mesh of the boiler (left) and its burner (right). ... 88 Figure 5.2 Predicted temperature contours inside the boiler under
air-coal and oxy-coal conditions. ... 95 Figure 5.3 Predicted velocity contours inside the boiler under
air-coal and oxy-air-coal conditions. ... 96 Figure 5.4 Predicted CO2 mole fraction profiles inside the boiler
under air-coal and oxy-coal conditions. ... 96 Figure 5.5 Predicted O2 mole fraction profiles inside the boiler
under air-coal and oxy-coal conditions. ... 97 Figure 5.6 Predicted heat transfer to different components under
air-coal and oxy-coal conditions. ... 98 Figure 5.7 A flowsheet of the virtually extended Didcot-A power
plant, including the original Didcot-A power generating unit,
an air separation unit and a CO2 compression unit. ... 99 Figure 5.8 A simplified thermal stage of a distillation column. ... 104 Figure 5.9 A simplified structure of the condenser/reboiler
between the high pressure and the low pressure columns. ... 105 Figure 5.10 Predicted evaporative heat transfer for the air-coal
and oxy-coal cases. ... 116 Figure 5.11 Predicted steam generation for the air-coal and
oxy-coal cases. ... 116 Figure 5.12 Predicted total steam generation for the air-coal and
oxy-coal cases... 117 Figure 5.13 Predicted radiative heat transfer for the air-coal and
oxy-coal cases. ... 118 Figure 5.14 Predicted convective heat transfer for the air-coal and
oxy-coal cases. ... 118 Figure 5.15 Predicted steam temperatures at the inlet/outlet of the
heat exchangers. ... 119 Figure 6.1 Part of the predicted temperature contours inside the
boiler. ... 136 Figure 6.2 ROMs for oxy-coal combustion of the boiler. ... 139 Figure 6.3 ROMs for the air-coal combustion of the boiler. ... 139 Figure 6.4 The predicted evaporative heat as a function of oxygen
Figure 6.5 The predicted steam generation as a function of the
oxygen concentration. ... 144 Figure 6.6 The predicted steam generation as a function of the
oxygen concentration. ... 144 Figure 6.7 The predicted radiative heat transfer as a function of
the oxygen concentration. ... 145 Figure 6.8 The predicted convective heat transfer to the
water/steam cycle as a function of oxygen concentration. ... 146 Figure 6.9 Predicted steam temperatures at the inlet/outlet of the
super heat components at 500MWe operation. ... 147
Figure 6.10 Predicted steam temperatures at the inlet/outlet of the super heat components at 400MWe operation. ... 147
Nomenclature
Abbreviations
ASU air separation unit CAD computer-aided design CBK carbon burnout kinetics CCS carbon capture and storage CFD computational fluid dynamics
Cov covariance
CPD chemical percolation devolatilisation CPU CO2 compression and purification unit CPU central processing unit
CTF combustion test facility DNS direct numerical simulation DOE design of experiments DOM discrete ordinates method DTM discrete transfer method
EDM eddy dissipation concept model
Eq equation
EDCM eddy dissipation concept model ESP electrostatic precipitator
FG-DVC functional group-depolymerisation vaporization cross-linking FGC flue gas condensation
FGD flue gas desulphurisation FGR flue gas recycle
FSCK full spectrum correlated-k HPC high pressure column
IGCC integrated gasification combined cycle LBL line-by-line
LES large eddy simulation LPC low pressure column MEA monoethanolamine MHT main heat exchanger
Mtoe million tonnes of oil equivalent
PACT pilot-scale advanced capture technology PDF probability density function
Plat1 superheater platen 1 Plat2 superheater platen 2 PRH primary reheater
PSA pressure swing adsorption
RANS reynolds-averaged navier-stokes PCA principal component analysis ROM reduced order model
RSM reynolds stress model RTE radiation transfer equation SCR selective catalytic reduction SGS sub-grid-scale
SNB statistical narrow band SSH secondary superheater SST shear stress transport UDF user defined function
WSGG weighted sum of gray gas
Latin alphabet
A effective heat transfer area 2
m
a absorption coefficient 1 m
b dimensional scaling coefficients -
p
C heat capacity J kg
F mass flow rate of the feed stream kg s
f number of degrees of freedom of the gas molecules -
f kernel function vector -
F response vector -
g gravity constant N kg
h mass specific enthalpy J kg
I radiation intensity 2
W m
k turbulent kinetic energy in Section 2.2.2 m s2 2
k chemical reaction rate in Section 2.3.1 mol s XR
K an empirical constant describing the pressure drop -
L mass flow rate of the liquid kg s
m mass fraction -
M mass holdup kg
n refractive index -
p pressure Pa
Q heat, energy J
Q total heat flow J s
r CO2 absorption rate mol s
R the correlation matrix -
s direction vector -
t time s
T temperature K
u velocity of the fluid in Chapter 2 m s
u mass specific internal energy in Chapter 5 J kg V mass flow rate of the vapour in Section 5.2.2.1 kg s
V volume flow rate in Section 5.2.2.2 3
m s
V volume in Sections 5.2.2.3, 5.2.2.4, 5.2.2.4 and 6.2.2 3
m
W adiabatic power J s
x mole fraction in Section 5.2.2.2 -
y height of the riser m
Greek alphabet
geometry coefficient of the furnace -
β regression coefficient vector -
Kronecker delta
correlation parameter -
wavelength in Chapter 2 m
index of the components in Chapter 5 -
rate of dissipation of turbulent kinetic energy m s2 2
adiabatic index of the gas
- Stefan-Boltzmann constant, 5.669X10-8 2 4 /
W m K in Section 2.2.1
square root of the process variance in Section 6.3.1 -
solid angle -
scattering phase function -
fugacity coefficient Pa
dynamic viscosity kg m s
density 3
kg m
specific dissipation rate= k in Chapter 2 1 s mass fraction in Chapter 6
mixed convection/radiation coefficient -Subscripts ad adiabatic flame av average circ circulation d steam drum eff effective evap evaporative
gen electricity generator
dc downcomer in inlet liq liquid mix mixture out outlet ox oxygen R riser ref reference
s steam
sat saturation
tfr heat transfer
vap vapour
WDC water in the downcomer w wall of the heat exchanger XR water/steam mixture
Chapter 1. Introduction and Motivation
In this chapter, the motivation for this investigation is introduced. The challenge of global warming and the necessity of using coal in the world energy mix are discussed in Section 1.1 and the use of coal and its impacts on the environment are analysed in Section 1.2. The solution for the continuous use of coal while achieving a low carbon emission, namely, the carbon capture and sequestration (CCS) technologies, are introduced in Section 1.3. A brief introduction on power generation system modelling techniques is presented in Section 1.4. Finally, the aims, novelties and the scope of this thesis are outlined in Section 1.5.
1.1 Energy consumption and the role of coal
Investment shows that the world energy consumption will drastically increase from 8,769 million tonnes of oil equivalent (Mtoe) in 1992 to 16,534 Mtoe in 2030 [1]. Further, there has been a worldwide upward in the demand of energy, with Brazil, Russia, India and China being the most likely biggest four economies in terms of energy consumption and demand over the next 2 decades, whose consumption levels of primary energy are even predicted as surpassing the OECD by 2030 [1]. Population growth has always been, and will remain, one of the key drivers of energy demand, along with economic and social development. The world population is expected to reach 8.1 billion in 2025 and 9.6 billion in 2050 [2], which leads to a more extensive demand on energy. Therefore, in order to maintain and improve people’s living standards, an increase in energy production is required.
Various types of fuels are used in the power producing industries to generate electricity: fossil fuels (coal, oil and natural gas), hydro, nuclear and renewables and Figure 1.1 describes the increasing trend of the demand on different fuels from 1988 to 2013. Figure 1.1 also reveals that the fossil fuels are the most depended energy sources and a more significant increase in the amount of consumption of coal is witnessed for the past two decades. Meanwhile, the use of coal always takes a remarkable role, which approximately occupies 30% of the total amount, in the whole mix.
Figure 1.1 World energy consumption by fuel [3].
Due to environmental policies, prices and technology developments, the demand on different fuels is always changing and Figure 1.2 shows the fuel use in the electricity supply in the UK from 1998 to 2013. It can be seen that the coal and gas contributions to electricity are significantly higher than those of other fuels. Moreover, since 2008, the use of gas has dropped gradually while the demand on coal has become relatively stable and even shown a mild rise.
Figure 1.2 Fuel used in electricity generation in the UK over the last 15 years [1]. 1988 88 10000 7000 4000 1000 13000 1993 1998 2003 2008 2013 World consumption Million Tonnes oil
equivalent M il li o n To nn e s o il e qu iv a le nt 80 60 40 20 0 1988 88 2001 2004 2007 2010 2013 100
Currently, fossil fuels are the most widely used sources for the electricity production. Considering the safety, economy, and abundance of the fossil fuels, coal comes first in accommodating human society’s demand. The reason is that the security, stability and capacity of supply are important actual issues that need to be considered:
(i) Although the Middle East countries have large amounts of oil reserve, the severe political and security environment of this region may become a barrier for the stable and continuous oil output; on the other hand, for some major developing countries (e.g. China and India) which are short of oil and gas but have considerably large amounts of coal reserve on which they can depend on and even export.
(ii) The clean energies, such as wind, solar and hydro, are environmental friendly and renewable. However, their capacities are too limited to meet all the demands and the stability of supply cannot be guaranteed since the weather and atmospheric conditions which they depend on always change. (iii) Nuclear power is an attractive alternative since it is considered the only kind of energy that has the potential to replace the fossil fuels for its high electricity producing capability. In addition, nuclear power is clean and does not bring in any unwanted gas emissions, such as CO2, SOX or NOX. However, the disastrous nuclear accidents (Chernobyl 1986 [2] and Fukushima, Daiichi’s 2011 [4] nuclear disasters) have warned people about the safety issues of the nuclear power. Following the Fukushima nuclear disaster, many countries have reshaped their nuclear development policies [5], e.g. Germany has decided to close all of its nuclear power stations by 2022 [5]. Fierce debates on nuclear power took place in Italy soon after the Fukushima nuclear disaster and its further nuclear development had been pending so far [5].
(iv) Biomass provides a new option for the energy mix. Biomass energy is mainly produced from plants, animals or other organic sources. It enjoys superiority in terms of sustainability due to the fact that burnt organic sources can release back CO2 and H2O into the air and a reproduction of plants and animals could be used to guarantee the circulation of energy generation. More importantly, the NOX or SOX emissions by burning biomass
are significantly lower than those of fossil fuels. However, depending on the biomasses can be expensive and its reproduction requires lots of land which may conflict with other demands for the use of land.
To summarise, in the foreseeable future, the demand on energy use will continuously increase and coal will still play a crucial role in meeting this demand.
1.2 Coal combustion and its impacts on the environment
1.2.1 Coal combustion in conventional power stations
The most important usage of coal is in electricity generation. The process of coal consumption in the traditional power plant can be seen in Figure 1.3.
1. Cooling tower 10. Steam governor 19. Superheater
2. Cooling water pump 11. High pressure turbine 20. Forced draught fan
3. Pylon 12. Deaerator 21. Reheater
4. Unit transformer 13. Feed heater 22. Air intake
5. Generator 14. Coal conveyor 23. Economiser
6. Low pressure turbine 15. Coal hopper 24. Air preheater 7. Boiler feed pump 16. Pulverise fuel mill 25. Precipitator
8. Condenser 17. Boiler drum 26. Included draught fan
9. Intermediate pressure turbine 18. Ash hopper 27. Chimney stack
Figure 1.3 Schematic of a coal fired sub-critical power plant.
In the furnace, the process water is converted to high pressure steam by the heat released from the coal combustion. The hot steam then goes through a set of steam turbines where the internal energy of the steam is turned into the mechanical energy of the turbines which drives the generator to produce electricity.
The heat transfer from the coal combustion to the heat exchangers is critical in the steam cycle. These heat exchangers, including the water walls, superheaters, reheaters and economisers, consist of several tube banks in order to enhance the effective area for heat transfer. The steam drum, which is located at the top of the boiler, is also an important component. Before entering the steam drum, the feed water passes through the economiser, which is a convective heat exchanger near the outlet of the furnace. Then the water in the steam drum goes down and into the tubes of the water wall, which surrounds the boiler. As the water passes through the water wall, the water is heated and becomes partially vaporised. This results in a decrease in the density of the water/steam along the water wall, thus the water/steam recirculates back into the steam drum. In the steam drum, the steam is separated from the water/steam mixture and is then passed to the superheaters to be further heated before entering the high pressure turbine. After driving the high pressure turbine, the steam recirculates back to the boiler to be reheated in the reheater, which is next to the superheaters. Then the reheated steam sequentially goes through the intermediate pressure turbine and low pressure turbine. The mechanical energy of the steam turbines is converted into electricity by a downstream generator. At the outlet of the low pressure turbine, the steam is condensed by the cooling water and then goes back to the economiser where another steam circle repeats.
Figure 1.4 Coal burner in a furnace in a power station [6].
As a dominant fuel used by the conventional power plants, coal is firstly ground in the mills to be very fine particles in order to enhance the combustion efficiency and then the pulverised coal is blown into the furnace with the carrying air via the burners. These burners are typically designed in order to reduce the pollutant formation and improve the combustion efficiency by bringing in strong turbulence/mixing between the coal particles
and the oxidant gas, which is achieved by adding swirled blades at the inlets of the burner (see Figure 1.4).
The furnace is the place where the coal combustion takes place and the chemical energy stored in the coal is converted into thermal energy which is transferred to the water wall and the superheaters mainly by radiation. As the high temperature flue gas passes through the superheaters, the gas temperature continues to decrease. When the flue gas reaches the economiser, the convection becomes the dominant form of heat transfer. Further, the flue gas contains some harmful acid gas, e.g. NOx and SOx, and therefore additional treatments for the acid gas removal are required before the flue gas is emitted into the atmosphere. Typical devices for the flue gas treatments are: electrostatic precipitator (ESP) to remove the particulate matter (soot or fly ash), the flue gas desulphurisation (FGD) equipment to remove the SOx and the selective catalytic reduction (SCR) unit to remove the NOx.
1.2.2 Impacts on the environment
The increase in the concentrations of the greenhouse gases is believed to be the reason for global warming and CO2 is recognised as the most important greenhouse gas. Global warming is an environmental phenomenon and the world’s average temperature has been continuously increasing since the industry revolution. The correlation between the CO2 emissions and the increase in temperature is simple: too much CO2 in the atmosphere obstructs the thermal radiation from the surface of the Earth to the outer space – like a thick quilt. Figure 1.5 shows a record of CO2 emissions since 1900 and it is clear that due to human activities, the global CO2 emissions have increased by more than 1000% since 1900. Consequently, the average atmospheric CO2 concentration level has increased by over 30% from about 296 ppm in 1900 to about 390 ppm in 2010 (see Figure 1.6).
Figure 1.5 Global CO2 emissions since 1900 [7].
Note: Concentration from two sources: measurements up to 1978 from Antarctic ice cores (blue), and direct atmospheric sampling at Hawaii since around 1960 (red).
Figure 1.6 Average atmospheric CO2 concentration since 1900 [7].
Figure 1.7 Sea level rise over the last 100 years [8].
One of the most concerned worries triggered by global warming is the melting of the huge glaciers around the world, which directly raises the sea level. It has been recorded that the sea level has risen by more than 150 mm
Em is s io ns from f os s il fue ls (G tCO 2 /y r) 8 10 6 4 2 0 1990 1920 1940 1960 1980 2000 2020 1990 1920 1940 1960 1980 2000 2020 A tm os ph e ric CO 2 c on c e ntra tio n (ppm ) 400 380 360 340 320 300 280 0 100 150 1900 1920 1940 1960 1980 2000 50 Cum ul a tiv e s e a l e v e l c ha ng e (mm )
over the last 100 years (see Figure 1.7). If we allow global warming to continue to develop without any control, then in several centuries that most of the land will be under the sea.
Facing this challenge, national and international efforts have been made to reduce greenhouse gas emissions. The Kyoto Protocol international agreement announced in 1997 that in order to commit countries who are members of the United Nations Framework on Climate Change (UNFCC) to reduce greenhouse gas emissions [9]. In Europe, short term and long term targets have been made regarding to greenhouse gas emission reduction: EU members have committed themselves to reducing greenhouse gas emissions by 20%, while increasing the share of renewables in the energy mix to 20% by 2020 [10]. In 2011, the EU confirmed a long term objective of reducing greenhouse gas emissions by 80-95% by 2050 compared to 1990 [10].
The UK, under the framework of UNFCC and the EU, aims to reduce 34% of the greenhouse gas emissions by 2020 and have a further reduction to 80% by 2050 compared to the 1990 level. Other major countries, such as China, United States, Canada, India and Brazil, have started their own program and policies to reduce greenhouse gas emissions [11].
1.3 Carbon capture technologies
The typical CO2 emission rate from a conventional coal-fired power plant can be as high as about 906 kg/MWh [12]. Therefore, coal-fired power plants are regarded as one of the most significant boosters to the atmospheric CO2 level. For example, from the top 50 dirtiest power plants in the USA, only less than 1% of the total number, produced 50% of all the USA’s vehicle carbon emissions [13]. Considering the importance of coal (see Section 1.1), coal still occupies a large share of the energy mix and will do so in the foreseeable future. Current environmental situations and government policies push energy extensive industries, especially coal-fired power plants, to develop new low carbon technologies.
Carbon Capture and Storage (CCS) represents a set of technologies that can capture more than 90% of the CO2 produced from burning fossil fuels in
electricity generation and other industrial processes, thus preventing the CO2 from being emitted to the atmosphere. The captured CO2 is liquefied and then transported by either pipe lines or ships to a suitable underground storage site which can be saline aquifers or depleted oilfields. Moreover, the stored CO2 can be utilised in other industrial sectors where pure CO2 is required. It has been acknowledged that the utilisation of CCS is a necessary way that people can keep fossil fuels in the world’s electricity supply mix while still meeting the greenhouse gas reduction requirements [14].
Generally, CCS technologies can be classified into three categories using different technique procedures, and these are pre-combustion, post-combustion and oxy-fuel post-combustion and the following part of this section provides a brief introduction to these three types of CCS technologies.
1.3.1 Pre-combustion
Figure 1.8 shows a simplified process diagram of the pre-combustion process. In pre-combustion technique, the CO2 is captured before the combustion process [15]. In the beginning, an air separation unit is used to produce pure O2, which is then mixed with a suitable amount of coal/fuel in a gasifier where a synthesis gas mainly consists of CO and H2 [15]. Further, the synthesis gas is passed to a reactor where the shift reaction with water takes place so that a mixture of CO2 and H2 is produced. Then, the CO2 can be captured, compressed and sequestered while the H2 can then be combusted in a gas turbine or a burner to generate thermal energy and more importantly the flue gas (mainly H2O) from combustion is 100% clean.
Air Separation Unit Air H2 Gasifier Coal Shift Reactor CO & H2 CO2 Capture CO2 & H2 CO2 Compression CO2 CO2 for storage Gas Turbine N2 O2
Figure 1.8 A simplified diagram for the pre-combustion process [16].
In the electricity generation sector, the pre-combustion technology can be used with carbon capture in an integrated gasification combined cycle (IGCC) power plant. A significant advantage of the IGCC power plant with carbon
capture is that its efficiency is about 7 - 9% higher compared to those of the oxy-fuel or post-combustion power plants [17]. However, the construction of IGCC power plants requires a high capital investment and this technology cannot be applied to the existing coal-fired power plants. Currently, pre-combustion technology is not yet fully commercialized. In the UK, several IGCC power plant projects are under consideration/construction, namely the Teesside Low Carbon Project (450 MW) with a CO2 capture ratio of 85%, the C.GEN North Killingholme Project (450 MW) in Yorkshire and the Don Valley Power Project (650 MW) in Yorkshire [18]. However, up to now, these projects have not been commissioned.
1.3.2 Post-combustion
Figure 1.9 represents a simplified process diagram of the post-combustion process where the CO2 capture process takes place after the combustion in the furnace [19]. The capture of CO2 could be achieved by allowing the flue gas to pass through some chemical solvent, which can be monoethanolamine (MEA) or methylenedioxyethylamphetamine (MDEA) or mixtures of them [19]. Then the CO2-rich solvent is heated to release the captured CO2 which is almost pure and ready for compression, and meanwhile the CO2-lean solvent is regenerated and recycled to the CO2 capture loop. In addition, just before the CO2 capture process, a gas cleaning process, where a flue gas desulphurisation (FGD) unit is employed, it is necessary to remove the SO2, which has an oxidative degradation effect on the MEA/MDEA solvent [20].
Furnace
Air & Coal CO2 & N2
Gas Cleaning CO2 Capture
CO2
CO2
Compression
CO2 for Storage
Pollutants Treated Gas
Figure 1.9 A simplified block diagram for the post-combustion process [16].
Post-combustion technology is a promising candidate for carbon capture and storage because it can be directly used to retrofit the existing coal-fired power plants. However, the integration of this technology would result in an efficiency penalty ( about 10% of the efficiency penalty with 90% of the CO2 captured [21] ) to the power plant because the regeneration of the lean solvent requires a steam extraction from the steam turbine to provide the necessary heat for the chemical reactions.
The world’s first commercial-scale post-combustion CCS project (SaskPower Integrated Carbon Capture and Storage Demonstration Project [22]) has been in operation in Canada. At full capacity, the post-combustion facility captures over 1 million metric tons of CO2 per year, reflecting a 90% CO2 capture ratio from a 139 MW coal-fired unit [22]. In July 2014, the world’s largest commercial post-combustion project (Petra Nova Project [23]) was announced in the USA. This project aims to install the post-combustion technology to the coal-fired W.A. Parish Generating Station to annually capture 1.4 million metric tons of CO2 from a 240 MW coal-fired facility, with a 90% CO2 capture ratio [23]. In the UK, a commercial post-combustion project, based on the Peterhead gas-fired power station in Aberdeenshire is under consideration [24] and the planning application is expected to be submitted in 2015.
1.3.3 Oxy-fuel combustion
Oxy-fuel combustion technology offers a viable low carbon pathway for the existing coal-fired power plants to enable CO2 capture and storage. The conventional coal-fired furnaces use air as the oxidant in the combustion process where the CO2 concentration in the flue gas is diluted by the nitrogen. In contrast, as is shown in Figure 1.10, the oxy-coal combustion in a furnace takes a mixture of oxygen and recycled flue gas as the oxidant gas in order to significantly increase the concentration of CO2 in the flue gases [25]. Generally, the purity (vol%) of the O2 used in the oxy-coal combustion is not less than 95% and for this purpose an ASU is employed [25]. The recycled flue gas is for the purpose of the flame temperature control and makes up the volume of the separated N2 to ensure there is sufficient gas to transfer the combusted heat to the heat exchangers. After the oxy-coal combustion, a flue gas, mainly consisting of H2O and CO2, is produced,
which is ready for compression and storage after a gas cleaning process where the SOx is removed [25].
Air Separation Unit Air N2 Furnace Coal O2 Gas Cleaning CO2 Compression Flue Gas
Recycle CO2 for Storage
Figure 1.10 A simplified block diagram for the oxy-fuel combustion process [16].
It should be noted that the use of an ASU in this technology brings in an energy penalty of about 10% [26] to the power generation system. The preferred method for the ASU is cryogenic distillation, since this technology currently is commercialised and is capable of producing a large amount of high purity oxygen compared to other oxygen separation technologies [27]. At the moment, oxy-fuel technology has not been commercialised and the Callide Oxy-fuel Project [28] is the only demonstration project of a oxy-fuel power plant in the world. The air-coal Callide-A power station in Queensland having a full load of 30 MW is retrofitted to an oxy-coal power plant.
In the UK, a oxy-fuel demonstration project with a gross output of 448 MW, named the White Rose CCS project [29], has been announced and the oxy-fuel power plant will be situated near to the Drax Power Station.
1.4 Power generation system modelling
System computer modelling techniques enable engineers to research and evaluate the power plant operation, optimisation and control policies so that the potential risk and cost of operating/constructing the power plant can be reduced.
In the modelling of a coal-fired power plant, accurate modelling of the boiler is important because even a small change in the combustion environment of the boiler may pose a significant impact on the overall performance of the
plant. In the boiler, the complex coal combustion process takes place and energy is released from the coal. The combustion process involves several steps: (i) the coal particle is preliminarily heated when entering the boiler; (ii) the moisture content in the coal is evaporated; (iii) as the coal particle absorbs more heat, the devolatilisation process takes place so that the volatile matters and tar is released; (iv) the char content left in the particle combusts as it is further heated. Correspondingly, in order to accurately model the combustion process in the boiler, the devolatilisation, volatile combustion and char combustion processes must be properly modelled. In addition, the strong turbulence in the boiler as the turbulence has an effect on the combustion process. Fortunately, computational fluid dynamics (CFD) is an important modelling technology in researching the combustion and fluid flow characteristics in the boiler and typically a commercial CFD code, named ANSYS FLUENT, can be used to cover these problems. ANSYS FLUENT employs the finite volume method to discretize the fluid domain enclosed by the boiler into a huge number of cells based on which the transport equations for the mass, momentum and energy balances are solved. The continuum gas phase is solved in an Eulerian frame [30] while the motion of the discrete coal particles is predicted in the form of a Larganian frame [30].
Apart from the boiler, a coal-fired power generation system contains many other components, such as the steam drum, steam turbines and the condenser. In addition, the air separation unit and the amine capture plant are involved in the whole system if carbon capture technologies are applied. It is impossible to wholly depend on CFD techniques to model all of these components due to the expensive computational resources and time required. Fortunately, process simulation techniques can cover this gap and there are several commercial process simulators available for this purpose, such as ASPEN Plus, gPROMS, PRO/II, DYNSIM, etc. Generally, process simulation employs simple mass and energy balance equations (zero or one-dimensional) to describe the modelled unit and numerous empirical parameters are employed. Therefore, the computational effort required is quite small compared to that employed in the CFD modelling.
In order to take advantages of both the CFD and process modelling techniques, integrated CFD and process co-simulation methods are becoming state-of-the-art in the research on the performance and integration of the power plant. It is clear that a 3D boiler CFD simulation usually takes a long time to obtain converged results while the process simulation accounts for the other components is much faster. Then if CFD and process modelling techniques are directly linked in such a way that the CFD simulation has to be performed at each of the operational conditions that are required in the plant process model. This approach is straightforward but requires an unacceptable amount of time for the CFD calculations to cover a whole range of operational conditions of a power plant [31]. Therefore, the efficient integration of CFD and process simulation techniques needs to be considered. Hence the reduced order model (ROM) technology provides a possible solution which is able to take the place of CFD models to very quickly obtain the necessary information (such as the heat flux to the water wall) to drive the process simulation [31].
1.5 Research aims, novelties and scope of the thesis
1.5.1 Research aims and novelties
Carbon Capture has been recognised as playing an important role in reducing the CO2 emissions from coal-fired power plants so that coal can be continued to be used in the energy mix. Both CFD modelling and process modelling techniques have been confirmed as important methods for investigating the application of the Carbon Capture technologies in the coal-fired power plants. Therefore, this research aims to develop a CFD and process co-simulation technique that can be depended upon to efficiently evaluate the operations of the power plants using carbon capture techniques. The novelties of this research are as follows:
i) More accurate reduced order models (ROM) have been developed to link the CFD to the whole plant process model.
ii) A new approach has been suggested for estimating the potential of retrofitting an existing power plant to oxy-firing.
iii) A feasible range of oxygen enrichments for the retrofitted power plant has been identified at different power loads.
1.5.1 Scope of the thesis
Concerning the technical issues discussed in Section 1.4, the research to be performed in this thesis can be divided into several milestones:
(i) In Chapter 2, a detailed literature review on oxy-coal system modelling techniques is presented, which involves CFD modelling and process simulation techniques. In the following Chapter 3, the experimental facility and data that are required for the model set up and validation in the thesis are summarised.
(ii) In Chapter 4, a set of combined CFD and process simulations is performed on an experimental facility, which involves a 250 kWth coal combustion furnace and a MEA based CO2 capture plant. The CFD techniques are employed to solve the turbulence, chemical reactions, and heat transfer in the coal combustion furnace while the process modelling techniques are used to account for the modelling of the CO2 capture plant. Then the reduced order models based on the CFD simulation results of the furnace are linked to the process model for CO2 capture plant.
(iii) In Chapter 5, the research objective is extended to the modelling of a large-scale coal firing power plant. A three dimensional CFD model for the utility boiler of this power plant and a process model for the whole power plant are developed. These efforts are the necessary preparations for developing a CFD and process co-simulation approach that can be employed to predict the operations of a power plant under both air-coal and oxy-coal firing conditions.
(iv) Based on the CFD and process models developed in Chapter 5, a new approach has been developed in Chapter 6 in order to estimating the potential of retrofitting an existing power plant to oxy-firing. The three dimensional CFD boiler model developed in Chapter 5 has been employed to simulate the complex coal combustion and heat transfer to the boiler heat exchangers under air-firing and oxy-firing conditions. Then, a set of reduced order models has been developed to link the CFD predictions to the whole plant process model, developed in Chapter 5, in order to simulate the
performance of the power plant under different load and oxygen enrichment conditions if retrofitted to oxy-firing. The reduced order models are generated based on the CFD simulations of the boiler using a non-linear Kriging interpolation method. With this new CFD-process co-simulation approach, the potential of retrofitting the Didcot-A power plant to oxy-coal firing is analysed.
Chapter 2. Literature Review
This chapter provides a detailed literature review on the modelling technologies with regard to the CO2 capture technologies that can be applied to the existing or new built coal firing power plants. The combustion process of the coal particles and modelling techniques are discussed in Section 2.1. The considerations of heat transfer and turbulence in CFD modelling are reviewed in Section 2.2. The process modelling approaches for the CO2 capture techniques that can be used in coal-fired power plants are discussed in Section 2.3. Finally, a brief summary about this chapter is provided in Section 2.4.
2.1 Coal combustion process modelling
Appropriate description and modelling of the combustion process of a single coal particle is important as it is fundamental for the modelling of the coal combustion in large scale boilers. The combustion process of a coal particle undergoes four major stages as described in Figure 2.1. In the evaporation process, the moisture content in the coal particle is evaporated; as the coal particle is further heated, the devolatilisation process takes place, where the volatile contents (light gases and tars) start to be released and react with the oxygen, which is known as volatile combustion. Then, as the temperature of the coal particle further increases, the char combustion process occurs, where the remaining char is oxidised at a lower rate compared to the devolatilisation and volatile combustion.
Evaporation Devolatilisation Volatile combustion Char combustion
Figure 2.1 Schematic of the combustion process of a coal particle [32].
It should be noted that the above description on the combustion process of a coal particle assumes that each stage takes place in a sequential order and this assumption is adopted in the current CFD codes. However, in fact, some of the stages may overlap.
2.1.1 Evaporation and devolatilisation
As the coal particle is heated by the surrounding gas quickly, the water evaporates fiercely once the temperature reaches the boiling point and the water escapes from the surface of the particle through many pores in the particle. During the evaporation process, the particle may shrink or break into smaller pieces, but this effect is currently not considered in the modelling techniques.
When the temperature increases further to about 600 K [33, 34], the light gases and tars, namely the volatile contents, begin to leave the particle through pores to the external gas phase and their subsequent oxidisation generates mainly CO2 and H2O. The physical structure of the coal particle changes significantly, which is related to the release of the volatile matter, and a swelling phenomenon can be observed [35]. The devolatilisation process is fundamentally affected by the coal type, temperature, pressure and the species of the surrounding gas [34]. After the devolatilisation process, the solid material remaining in the particle is the char, which has a porous structure. In fact, the structure and reactivity of the char is affected by the devolatilisation process [32, 36].
Clearly, the amount of volatile content released from devolatilisation varies for different coal types. Coal can be classified into three main categories, namely the lignite, bituminous and anthracite, according to their ages [32]. As the youngest coal, lignite is comparatively soft and mainly contains moisture and volatile matters with low fixed carbon, while the anthracite, as the oldest coal, is comparatively hard and mainly contains fixed carbon with little moisture and volatile matter [32]. The amount of volatile matter present in the bituminous coal lies between the other two types of coal [32]. In addition, it had been found that the amount of volatile matter released could be enhanced by a higher peak temperature and higher heating rates [37-39]. The amount of the volatile matter in the coal can be measured from a
drop-tube furnace with controls on the heating rate. A factor called the ‘high temperature volatiles yield ratio’ is usually employed to describe this enhancement by comparing the amount of the obtained volatiles to that measured from a standard proximate analysis [37].
The rate of devolatilisation can be modelled by a single-rate model [39] using a single Arrhenius formation where the devolatilisation rate is assumed to be proportional to the volatiles remaining in the particle. As a matter of fact, the volatile mater leaves the particle at various rates, thus the single-rate model may be insufficient to accusingle-rately describe the process. A more suitable solution with higher fidelity would be the two-competing rate model, which was developed by Kobayashi et al. [40]. The two-competing rate model relies on six parameters and is capable of modelling most coals, if the corresponding data for the coal is available. Silaen et al. [41] investigated different devolatilisation models as a part of a CFD code. They found that the two-competing rate model predicted a slower devolatilization rate than the single-rate model but produced a higher exit gas temperature and higher CO2 mass fractions. However, experiments were not performed. The Sandia National Laboratories [42] performed a number of experiments and found that the model constants used by Kobayashi et al. [40] could not give satisfying predictions on some coals, while the constants used by Ubhayakar et al. [43] appeared to increase the accuracy.
Network models, such as the chemical percolation devolatilisation (CPD) model [44], the functional group-depolymerisation vaporization cross-linking (FG-DVC) model [45] and FLASHCHAIN [46], can predict the devolatilisation rate and the yields of gases and tars under different heating rates if the structure parameters of the coal particle are available. Jones et al. [47] evaluated different devolatilisation models and concluded that the network models could provide satisfactory devolatilisation rates. William et al. [48] performed experiments on a drop-tube furnace for a range of coals and the experimental results were compared to the predictions from the CPD, FG-DVC and FLASHCHAIN models and the predictions on the volatile yields were in generally good agreement with the experimental data, although these models predicted slightly different results.
Rastko et al. [49] implemented the single rate, two competing rates, CPD and FG-DVC model as a part of a CFD code in order to numerically determine the ignition point of a bituminous coal in a laboratory ignition test facility under air and oxygen enriched environments. The predictions suggested that the network models (CPD and FG-DVC) provide more accurate results compared to the single rate and the two competing rates models and the best performance was achieved by the FG-DVC model. However, the authors indicated that the use of the FG-DVC model would require much more computational resources, since the additional transport equations for the volatile species need to be solved. The results also revealed that the devolatilisation models, which were originally developed for conventional air combustion, can be applied to oxygen enriched combustion conditions. Moreover, Shaddix et al. [50] found that the switching to an oxygen enriched combustion environment has little impact on the devolatilisation process if the combustion temperatures are kept the same.
2.1.2 Volatile combustion
The volatile matters are released from coal particles mainly contain CO, CO2, H2O and many hydrocarbons [36]. The volatiles then react with the surrounding oxidiser gas to produce CO2 and H2O with numerous intermediate products. Therefore, the accurate description of the volatile combustion process involves a large number of intermediate reactions and species [32], which would pose a significant challenge for the CFD modelling as numerous chemical mechanisms and transport equations need to be solved. A popular solution and simplification is a global mechanism, which assumes the reaction rates to be very fast and greatly reduces the number of reactions and species. The global mechanism assumes the volatile matters to be a single material CxHyOz and its oxidisation can be represented as the following two step reaction [51]:
x y z 2 2
y 2x+y-2z
C H O +αO xCO H O, where α
2 4 (2.1) 2 2 1 CO O CO 2 (2.2)
In this global mechanism, an intermediate species ‘CO’ is introduced to describe the char oxidation occurring on the particle surface and the
intermediate CO is further oxidised to CO2. In fact, the complex combustion process is highly determined by the turbulent mixing, which affects the kinetic rates of the reactions.
For the laminar flames, the reactions in the combustion gas phase can be described by the kinetic rates. However, in a pulverized coal combustion furnace, the volatile combustion is significantly coupled with the strong turbulence. It is known that strong turbulence may greatly enhance the reaction rates [52] and this makes the global mechanism theoretically applicable. In fact, the proper selection of an approach to model the volatile combustion involves several aspects that need to be considered: accuracy, the ability to describe the chemical reactions and the computational requirement.
The eddy dissipation model (EDM) [53] relates the reaction rates with the turbulence level. However, this model ignores the chemical kinetics, thus it does not account for the intermediate species and can be only used with the global mechanism. The EDM has been widely used for modelling the pulverized coal combustion process [54-57]. The eddy dissipation concept model (EDCM) [58], as an expansion of the EDM, takes the detailed chemical kinetics into consideration and therefore this model can describe the intermediate reactions and products. However, this requires the solution of the transport equations for each species, and thus the demand on the computational resources is significantly higher.
The laminar flamelet model [59] treats the turbulent flame as a set of laminar flamelet regions. This model considers a larger number of chemical reactions and intermediate species. The characteristic of this model is that the density, species fractions and the temperature profile near the flamelet are described by the mixture fraction and the scalar dissipation rate [60]. However, the scalar dissipation rate needs to be modelled separately. Combined with a PDF, the flamelet model can be used to model the turbulent flames.
The application of the probability density functions (PDF) [61] provides a promising option to address the chemical kinetics in the combustion flows. The significance of this approach is that the chemical source term can be
easily expressed. However, this approach is computationally expensive. The mathematic formation of the PDF needs to
